The Internet, by its design, is basically an anonymous medium. The Internet is a group of networked computers with each of the computers being assigned an IP addresses. The IP address is a set of four numbers that are used in routing communications between the computers. While the IP address provides useful routing information, the IP address really does not provide any information about the actual user. Consequently, when a user visits a web site, the user will have just its IP address revealed to the web site and the web site cannot directly detect anything else about the user.
Despite the anonymity of the Internet, many sites on the Internet strive to learn something about users that visit their site. These sites believe that by personalizing the site to the users, the users have a more enjoyable experience and are more likely to be a repeat user. By personalizing the site to the user, sites also hope that they can increase the effectiveness of the site, whether it be increased sales, increased revenue from advertising, or simply increased traffic. The personalization can take many forms, including providing a special greeting which mentions the user by their name, customized settings such as a MyPage, advertising that is targeted to the interests of the user, and automatically logging the user into the site.
One common way that many sites perform personalization will now be described with reference to FIG. 1. A user 5 goes through the Internet 10 to a web site 15. During the course of a visit at the web site 15, the user 5 provides or reveals some useful information about the user 5. The information may information input by the user, such as the user's name, address, or email address, or may be information that is more indirectly detected by the web site 15, such as the type of information that is of interest to that user that is estimated by the requested pages. At this time, the web site 15 now has the IP address of the user and some information about the user. The web site 15 could construct a database associating that information with that IP address. A problem with this approach is that many IP addresses are dynamically assigned to users so the web site cannot know for sure if the same user 5 is associated with any given IP address.
To provide some correlation between information about a user and a user's computer 5, many web sites 15 employ cookies. A cookie is a small data file that is sent from the web site 15 and stored on the user's computer 5. Most computers have browsers that allow a user to control the storage of cookies on their computer 5 and give several options so a user can choose to prevent any cookie from being stored, to accept all cookies, or to accept cookies only after being warned that a cookie is about to be written to the computer 5. The web site 15 can therefore create a cookie that provides some identification of the user and then store the cookie on the user's computer 5. The next time that the user visits the web site 15, the web site 15 reads the cookie from the user's computer 5 and then retrieves the information about the user from the web site's database in order to provide some personalization. The web site 15 can therefore remember aspects of the user even if the user is assigned a different IP address the next time the user visits the site 15.
As mentioned above, cookies allow web sites to internally correlate a user with some information and typically include some coded information, such as an id number, a reference domain name, and a series of numbers. The information within a cookie is typically useless to another web site. Thus, a second web site 20 first of all probably would not even know that the web site 15 stored a cookie on the user's computer 5. Even if the web site 20 did know about the cookie and read that cookie, the cookie would not by itself give the web site 20 any useful information on the user 5. Consequently, the second web site 20 needs to obtain or detect information about the user 5, create its own database, and place its own cookie on the user's computer 5.
In addition to requiring duplication of effort at the various web sites, the need for each web site to store its own cookies on a user's computer 5 also requires duplication of effort by the user. Each time a user visits a new web site, the user provides information to the site and/or goes through a process whereby the site can detect information about the user. For example, the user may need to register at the site and/or request a set of pages before the web site can learn information about the user. Unfortunately, the user cannot easily provide such information to sites but instead must go through this trouble with each new site.
Even with the best profiling technology, a web site can learn only a limited amount of information about a user. A web site has a relatively brief encounter with a user and from this brief encounter the site tries to learn as much as possible about the user. A single visit to a site can provide some clues about the user, such as the user's interests. A single visit, however, is just a snap shot in time and may not accurately reflect the true interests of the user. The information may prove to be inaccurate for many reasons, such as the user was searching for a friend, another person was using the user's computer, the user's interests changed over time, or the web site simply guessed wrong in estimating the information. A need therefore exists for ways in which to more accurately obtain useful and accurate information on users.